116 THE PHYSICAL GEOGRAPHY OF THE SEA. 



CHAPTER V. 



EED FOGS AND SEA DUST. 



Where found, ^ 266.— Tallies on the Wind, 272.— Where taken up, 278.— Humboldt's 

 Description, 282. — Questions to be answered, 284. — What Effects the Deserts have 

 upon the General Circulation of the Air, 286. — Information derived from Sea Dust, 

 288.— Limits of Trade-winds, 289.— Breadth of Calm Belts, 290. 



266. Seamen tell us of "red fogs" which they sometimes en- 

 counter, especially in the vicinity of the Cape de Verd Islands. 

 In other parts of the sea also they meet showers of dust. What 

 these showers precipitate in the Mediterranean is called " sirocco 

 dust," and in other parts " African dust," because the winds which 

 accompany them are supposed to come from the Sirocco desert, or 

 some other parched land of the continent of Africa. It is of a 

 "brick-red or cinnamon color, and it sometimes comes down in such 

 quantities as to cover the sails and rigging, though the vessel may 

 be hundreds of miles from the land. 



267. Now the patient reader, who has had the heart to follow 

 me in the preceding chapter around with "the wind in bis cir- 

 cuits," will perceive that proof is yet wanting to establish it as a 

 fact that the northeast and southeast trades, after meeting and ris- 

 ing up in the equatorial calms, do cross over and take tKe paths 

 represented by C and G, Plate I. 



268. Statements, and reasons, and arguments enough have al- 

 ready been made and adduced to make it highly probable, accord- 

 ing to human reasoning, that such is the case ; and though the 

 theoretical deductions showing such to be the case be never so 

 plausible, positive proof that they are true can not fail to be re- 

 ceived with delight and satisfaction. 



269. Were it possible to take a portion of this air, representing, 

 as it travels along with the southeast trades, the general course of 

 atmospherical circulation, and to put a tally on it by which we 

 could follow it in its circuits and always recognize it, then we 



